Content uploaded by Thomas C Williams
Author content
All content in this area was uploaded by Thomas C Williams on Jun 24, 2021
Content may be subject to copyright.
Page Proof Instructions and Queries
Greetings, and thank you for publishing with SAGE. We have prepared this page proof for your review. Please respond to each of the
below queries by digitally marking this PDF using Adobe Reader.
Click “Comment” in the upper right corner of Adobe Reader to access the mark-up tools as follows:
For textual edits, please use the “Annotations” tools.
Please refrain from using the two tools crossed out below,
as data loss can occur when using these tools.
For formatting requests, questions, or other
complicated changes, please insert a comment using
“Drawing Markups.”
Detailed annotation guidelines can be viewed at: http://www.sagepub.com/repository/binaries/pdfs/AnnotationGuidelines.pdf
Adobe Reader can be downloaded (free) at: http://www.adobe.com/products/reader.html.
No. Query
Journal title: Journal of Human Lactation
Article Number: 632203
1 Please confirm that all author information, including names, affiliations, sequence, and contact details, is correct.
2 Please review the entire document for typographical errors, mathematical errors, and any other necessary corrections; check headings, tables, and figures.
3 Please confirm you have reviewed this proof to your satisfaction and understand this is your final opportunity for review prior to publication.
4 Please check all citations in this paragraph, as the number of cites do not always match the number of studies mentioned in the text.
5 Six citations are given. Please confirm the correct citations
6 Please confirm that the Funding and Conflict of Interest statements are accurate.
7 Please provide editor(s) if applicable, page range, and publisher
8 Please provide the city of publication
9 Please provide the date of the meeting and the location
10 Please provide the volume number and confirm the page range
11 Please provide the date this document was accessed
632203JHL
Journal of Human Lactation
1 –9
© The Author(s) 2016
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0890334416632203
jhl.sagepub.com
Review
Donor human milk (DHM) is used in neonatal intensive care
units (NICUs) for the feeding of preterm infants when own
mother’s milk (OMM) is not available or insufficient. A
recent Cochrane review1 showed that in preterm and low
birth weight infants, feeding with formula compared with
DHM results in a higher risk of developing necrotizing
enterocolitis (NEC). As the incidence of NEC increases in
relation to the other complications of preterm birth,2 there is
growing interest worldwide in the use of DHM.
Currently, it is estimated that there are about 500 human
milk banks (HMBs) in existence in over 37 countries.3 In
addition, the number of HMBs is known to be growing
in countries with large populations such as India,4 and the
first HMB in Russia was recently established in Moscow.5
Donor human milk is currently recommended by the
World Health Organization (WHO),6 the American Academy of
Pediatrics (AAP),7 and the European Society for Paediatric
Gastroenterology Hepatology and Nutrition (ESPGHAN),8 as
the preferred alternative to OMM if this is not available for
low birth weight (WHO) or preterm (AAP, ESPGHAN)
infants. In the United States, the proportion of NICUs using
DHM increased from 25% in 2007 to 45% in 2011.9
Despite this, there remain many neonatal units that do not
use DHM, for a variety of reasons, including cost, uncer-
tainty about the evidence base for its use, and parental prefer-
ences.10 A 2014 survey of level 3 and 4 NICUs in the United
States10 showed that that 41% of respondents did not use
DHM for their patients. Similarly, a survey of special care
baby units, local neonatal units, and NICUs in the United
Kingdom,11 also carried out in 2014, showed that 39% of
respondents did not initiate infants on DHM.
If the use of DHM continues to increase, a key question is
how this may affect maternal breastfeeding rates. A national
XXX10.1177/0890334416632203Journal of Human LactationWilliams et al
research-article2016
1Simpson Centre for Reproductive Health, Edinburgh, UK
2Centre for Population Health Sciences, University of Edinburgh, UK
3Royal Hospital for Sick Children, Yorkhill, Glasgow, UK
4Newcastle Hospitals NHS Foundation Trust, Newcastle upon Tyne, UK
Date submitted: April 20, 2015; Date accepted: January 18, 2016.
Corresponding Author:
Thomas Williams, MRCP, Neonatal Unit, Simpson Centre for
Reproductive Health, Royal Infirmary of Edinburgh, 51 Little France
Crescent, Edinburgh EH16 4SA, UK.
Email: Thomas.Williams@cantab.net
Use of Donor Human Milk and Maternal
Breastfeeding Rates: A Systematic Review
Thomas Williams, MRCP1, Harish Nair, MD, PhD2,
Judith Simpson, MD3, and Nicholas Embleton, MD4[AQ: 1][AQ: 2][AQ: 3]
Abstract
The number of human milk banks is growing worldwide. The introduction of donor human milk (DHM) to neonatal
units has been advocated as a strategy to promote maternal breastfeeding. However, concern has been raised that the
introduction of DHM may actually lead to a decrease in maternal breastfeeding. To address this question, we conducted
a systematic literature review of studies that assessed maternal breastfeeding rates before and after the introduction of
DHM. We searched 7 electronic databases, carried out citation tracking, and contacted experts in the field. Where data
for breastfeeding rates before and after the introduction of DHM were directly comparable, a relative risk was calculated.
Our search identified 286 studies, of which 10 met the inclusion criteria. Definitions of patient populations and study
outcomes varied, limiting meaningful comparison. Where possible, relative risks (RR) were calculated on aggregated data.
The introduction of DHM had a significant positive impact on any breastfeeding on discharge (RR, 1.19; 95% confidence
interval [CI], 1.06-1.35; P = .005) but none on exclusive maternal breastfeeding on discharge (RR, 1.12; 95% CI, 0.91-1.40;
P = .27) or on exclusive administration of own mother’s milk (OMM) days 1 to 28 of life (RR, 1.08; 95% CI, 0.78-1.49;
P = .65). A single-center study demonstrated a significant decrease in the percentage of feeds that were OMM after the
introduction of DHM. In conclusion, the available data demonstrate some evidence of positive and negative effects on
measures of maternal breastfeeding when DHM is introduced to a neonatal unit.
Keywords
breastfeeding, donor human milk, milk banks
2 Journal of Human Lactation
survey in Italy showed that neonatal units associated with an
HMB have higher rates of maternal breastfeeding on dis-
charge.12 Using these data, these authors argued that the
introduction of DHM may serve to extend a culture of breast-
feeding. Similarly, others have argued that DHM should be
considered a supportive measure to mothers expressing milk
for their preterm infants and have used it as part of package
of measures to try to increase maternal breastfeeding rates13
and promote a culture of using only human milk14 on NICUs.
However, anecdotally, concerns have been raised that the
introduction of DHM to a NICU may in fact discourage
maternal breastfeeding.13,14 In addition, the authors of one
study have shown that promoting DHM can lead to an unin-
tended decrease in the use of OMM, perhaps by providing an
“acceptable alternative” to the initiation and maintenance of
lactation.15 There is thus uncertainty as to whether the further
introduction of DHM will affect either positively or nega-
tively on maternal breastfeeding rates in NICUs.
Two large trials in North America are currently addressing
whether there are clinical benefits to infants of using DHM
compared with formula.16,17 However, because these trials are
both blinded, impacts on health professional or maternal behav-
iors will not be fully determined. Thus, the aim of this review
was to strengthen the evidence base for the use of DHM and to
determine the effects of DHM provision on measures of OMM
use during admission and on breastfeeding rates at discharge.
We addressed the following research question: in mothers with
an infant admitted to a neonatal unit (population), what are the
effects of using DHM (intervention) versus formula milk (com-
parison) on maternal breastfeeding rates in, and on discharge
from, the NICU (outcome)? Given the complexity of DHM as
an intervention, we anticipated there might be relatively few
randomized controlled trials and that cluster trials and/or obser-
vational studies would require research synthesis.
Methods
Searches
This review and the manuscript reporting it was prepared
according to the PRISMA guidelines,18 and the completed
PRISMA checklist is available in Supplementary Appendix
S1, available online. We carried out a systematic literature
review in October 2014 using the following databases:
Medline,19 Embase,20 and Global Health21 (all using the OVID
interface)22; The Cochrane Library23; CINAHL24; Global
Health Library25; and Current Controlled Trials.26 Search
terms were generated using MeSH and Emtree terms relating
to breast milk, infant formula, milk banks, milk donation, and
neonatal units, with input from a medical librarian. A complete
list of search terms, formatted for each database, is available
within the study protocol in Supplementary Appendix S2. The
review is registered on PROSPERO27 (CRD42014013162).
Databases were searched from 1946 onward. Only studies
with abstracts published in the Latin alphabet were reviewed,
and these were translated if necessary by one of the authors
(T.W.). We conducted reference searches of the studies that
met the inclusion criteria, carried out citation tracking of
these studies via Google Scholar,28 and contacted experts in
the field in North America, Europe, and Australia to identify
further relevant studies. Two reviewers (T.W. and J.S.) inde-
pendently assessed the articles identified in the screening
search using the inclusion and exclusion criteria.
Inclusion/Exclusion Criteria
Studies were included if (1) the study was original research,
(2) the study was a controlled trial with participants allocated
randomly or an observational trial examining the impact on
maternal OMM provision or breastfeeding rates pre- and pos-
tintroduction of DHM to a neonatal unit, (3) the study popula-
tion was infants admitted to a neonatal unit, (4) the study
specifically compared enteral feeding with DHM versus for-
mula, and (5) the study provided quantitative data on maternal
breastfeeding rates during the admission or on discharge.
Studies were excluded if the patient population included
infants in postnatal or pediatric wards or did not compare
donor breast milk directly with formula. Study types that
were excluded were (1) case reports or opinion pieces with-
out primary data or (2) qualitative studies that did not pro-
vide data on the proportion of mothers’ breastfeeding during
the admission or on discharge.
Data Extraction, Assessment of Study Quality,
and Risk of Bias
The following data were extracted from the studies meeting
the inclusion criteria: authors, study setting and country
where it took place, research question/study aims, definition
of patient population, outcome measure, study sample size,
rates of breast milk use prior to introduction of donor milk to
a unit, and rates of breast milk use after the introduction of
donor milk. Where data were given for breastfeeding rates on
discharge, it was noted where this was on discharge from and
the definition of the time period used (eg, within 48 hours of
discharge). Data were entered onto Microsoft Excel
(Microsoft Corporation, Redmond, WA, USA).
To assess the risk of bias within each individual study, we
applied principles from the Cochrane Collaboration and the
Working Group for Grading of Recommendations Assessment,
Development and Evaluation.29 Modifying a scoring system
used previously by one of us,30 we assessed the quality of each
study as being high, moderate, or low, according to study
design, sample size, quality of the control group, calculation of
an odds ratio/relative risk, confounding factors, and the geo-
graphical spread of studies. Details of the scoring system can
be found in Supplementary Appendix S3.
To assess the risk of bias across studies, we noted
whether or not each study had been published in a peer-
reviewed journal. We contacted the principal authors of
Williams et al 3
each included study to ascertain if they could share any
unpublished data that might influence the cumulative evi-
dence available. Finally, we contacted experts in the field to
ensure there were no large data sets that were unavailable
due to publication bias.
Data Analysis
Where data for breastfeeding rates after the introduction of
DHM were directly comparable between studies, the num-
bers of infants in each group were aggregated and a relative
risk (RR) with a 95% confidence interval (CI) was calcu-
lated.31 The exposure for these calculations was the intro-
duction of DHM to a neonatal unit. Where data were not
comparable between studies, the outcomes before and after
the introduction of DHM were extracted, and it was noted
whether a summary measure had been calculated.
Results
Searches
Our database search yielded 374 records, and consultation
with experts in the field identified 4 further studies. Citation
tracking of studies that met the inclusion criteria yielded 25
additional records, and after excluding duplicates, a total of
286 studies were screened. Fourteen of these studies were
selected for full-text review, of which 10 studies met the
inclusion criteria (Figure 1). Of the remaining 4, 1 was
excluded as it duplicated data from an included study,32 1
was not based in a neonatal unit,33 1 did not compare breast-
feeding rates before and after the introduction of DHM,34
and 1 provided no quantitative data on breastfeeding rates.35
Six of the included studies were based in the United
States,13,14,36-38 2 in Spain,39,40 and 1 study in the United
Kingdom41 and Australia.42 All the studies were published
since 2008 but included data on infants born between 2001
and 2014. Table 1 provides a summary of the study
characteristics.
Quality Assessment and Risk of Bias
The assessment of study quality is shown in Table 2. One of
the studies was judged to be of high quality,39 7[AQ: 4] of
the studies were assessed to be of moderate quality,14,36-38,41
and the remaining 213,42 to be of low quality. Only 1 study
was prospective and interventional,14 and only 2 included
more than one hospital site.13 Two studies included DHM as
part of a bundle of measures designed to increase maternal
breastfeeding rates.13,14 Six were published in peer-reviewed
journals,13,14,36,37,40 and 4 were conference abstracts.38,41,42
None of the contacted authors of the included studies shared
unpublished data to contribute to our analysis. Consultation
with experts in the field did not reveal any large unpublished
data series relevant to this review.
Definitions
There was substantial heterogeneity in the definition of the
patient population in the included studies. One study looked
at infants born at < 30 weeks’ gestation,42 2 studies examined
infants born at < 32 weeks’ gestation or with a birth weight
(BW) of < 1.5 kg,13,40 5 studies used a BW < 1.5 kg as the
inclusion criteria,36-38,41,43 1 study used BW < 2 kg,14 and 1
study used BW < 1 kg.39 Outcome definitions were similarly
heterogeneous and were comparable in 4 studies for any
breastfeeding on discharge14,39,40,43 and in 2 studies for exclu-
sive breastfeeding on discharge40,42 and exclusive adminis-
tration of OMM days 1 to 28.37,40 Only 1 study40 defined a
time period before discharge for the receipt of breast milk
(48 hours), and none of the studies defined how that breast
milk was given on discharge. Five[AQ: 5] studies did not
document whether all infants or only surviving infants were
used as the denominator for measures of maternal breast-
feeding,13,14,36,38,42,43 3 excluded infants who died from their
analysis,37,40,41 and 1 study included these in the denomina-
tor.39 When performing calculations, the denominators used
were those given by the authors, and no adjustments were
made for the infants who died, as these numbers were small.
Studies also varied in how DHM had been introduced to a
neonatal unit. Three studies looked at changes in the admin-
istration of OMM after the introduction of a milk bank to a
neonatal unit.40-42 Two studies examined whether there was a
change in practice after the introduction of DHM as part of a
bundle aimed to increase the use of human milk.13,14 One
examined changes in practice after a new policy specifying
use of DHM when not enough OMM was available,37 and the
remaining 4 examined changes in practice after the introduc-
tion of DHM to a neonatal unit.36,38,39,43
Effects of Introduction of DHM on Maternal
Breastfeeding Rates
Two studies40,42 examined the effect of the introduction of
DHM on exclusive maternal breastfeeding rates on dis-
charge. One of these provided no definition of “exclusive
breastfeeding” on discharge, and the studies included 2 dif-
ferent patient population groups (born at < 30 weeks42 vs
born at < 32 weeks or BW < 1.5 kg).40 Aggregating the data
showed no significant difference between the 2 groups, with
an RR of 1.12 (95% CI, 0.91-1.40; P = .27) of breastfeeding
on discharge after the introduction of DHM.
Four studies14,39,40,43 provided data on infants receiving
any breastfeeding on discharge after the introduction of
DHM. In one of these studies,14 DHM was introduced as part
of a program aimed to increase the volume of human milk
given to infants born at less than 2 kg. No significant differ-
ence (P = .09) was found in infants receiving any breast
feeds on discharge after the introduction of the program. No
formal definition was given of “any breastfeeding on dis-
charge” in this study. Another study43 found a significant
4 Journal of Human Lactation
increase (P = .02) in any breastfeeding on discharge after the
introduction of DHM milk to a neonatal unit. Patient popula-
tion groups differed between the 4 study groups (BW < 2
kg,14 < 1.5 kg,43 < 1 kg39 and born at < 32 weeks or BW < 1.5
kg).40 Aggregating the data for the 4 studies, a significant
difference was found between the 2 groups (RR, 1.19; 95%
CI, 1.06-1.35; P = .005), showing an increase in maternal
breastfeeding after the introduction of DHM.
Two studies37,40 examined the effect of the introduction of
DHM on the exclusive administration of OMM in the first 28
days of life. One used a patient population of infants born at <
32 weeks,37 and another looked at infants born at < 32 weeks or
with a BW < 1.5 kg.40 In the second study, there was a reduc-
tion (from 40% to 13%) in the percentage of infants receiving
exclusive OMM. According to the authors, this was because
after the introduction of DHM, it was used when there was not
enough milk from the infants’ own mothers, whereas prior to
the introduction of DHM, infants were fed by parenteral nutri-
tion the first days of their lives to avoid infant formula.
Aggregating the data, no significant difference was found
between the 2 groups (RR, 1.08; 95% CI, 0.78-1.49; P = .65).
All the data above are shown in Supplementary Appendix S4.
Single studies provided data on a number of variables
related to the use of OMM after the introduction of DHM
Figure 1. PRISMA Flowchart.
Abbreviations: BF, breastfeeding; DEBM, donor-expressed breast milk.
5
Table 1. Summary of Study Characteristics.
Authors
Study
Setting
(Country) Research Question/Study Aims
Definition
of Patient
Population Outcome
Study
Sample
Size Main Findings
Beasmore et al41 NICU
(United
Kingdom)
Did the introduction of an HMB change OMM and
formula milk usage during the establishment of
enteral feeding?
BW < 1.5 kg % Exclusive OMM until
full enteral feeds
achieved
122 65% pre vs 70% post,
P = .51
Bishop et al36 NICU
(United
States)
To assess the influence of DHM on the incidence of
NEC and the amount of OMM use
BW < 1.5 kg % of feeds that contained
> 50% OMM up to 34
weeks CGA
331 51% pre vs 54% post,
P = .95
Delfosse et al13 Level 4 NICU
(United
States)
To determine acceptance of DHM for feeding
preterm infants and whether offering DHM alters
OMM feeding
Born at <
32 weeks
or BW <
1.5 kg
% OMM given (volume)
days 1 to 14 of life
650 63% at start of
intervention vs 60% at
end of intervention,
no P value calculated
Esquerra-Zwiers
et al15
Level 4 NICU
(United
States)
To evaluate the impact of a DHM program on OMM
and formula feedings
BW < 1.5 kg % Feeds that were OMM
days 1 to 14 of life
265 85% pre vs 68% post,
P < .01
% Feeds that were OMM
days 1 to 28 of life
265 71% pre vs 61% post,
P = .04
Kok et al42 Neonatal unit
(Australia)
The effects of the introduction of an HMB on the
feeding of preterm infants on discharge
Born at < 30
weeks
Exclusive BF on discharge
from neonatal unit (not
specified further)
155 53% pre vs 64% post,
no P value calculated
Marinelli et al37 Level 4 NICU
(United
States)
To compare enteral intake type in preterm infants
before vs after establishing a DHM policy
BW < 1.5 kg % OMM given (volume)
days 1 to 28
154 66% pre vs 70% post,
no P value calculated
Exclusive administration
of OMM days 1 to 28
154 38% pre vs 55% post,
no P value calculated
Montgomery et al14 Level 3 NICU
(United
States)
To assess the effects of a program designed to
improve human milk availability for preterm infants
on breast milk use and feeding-related outcomes
BW < 2 kg Receiving any BF on
discharge home (not
specified further)
245 44% pre vs 53% post,
P = .09
Parker et al43 Level 3 NICU
(United
States)
To determine whether rates of consumption of
OMM at discharge home changed in the 2 years
pre- and postimplementation of a DHM program
BW < 1.5 kg Any BF on discharge
from hospital (not
specified further)
154 43% pre vs. 65% post,
P = .02
Utrera Torres
et al40
Neonatal unit
(Spain)
To assess the impact that opening an HMB had on
the proportion of infants breastfeeding at discharge
and other practices related to feeding
Born at <
32 weeks
or BW <
1.5 kg
Exclusive BF on discharge
from hospital (within 48
hours of discharge)
104 54% pre vs 56% post,
P = .87
Any BF on discharge
from hospital (within 48
hours of discharge)
104 86% pre vs 78% post,
P = .27
Exclusive administration
of OMM days 1 to 28
104 40% pre vs 13% post,
no P value calculated
Verd et al39 NICUs
(Spain)
To assess the impact of an exclusive human milk diet
to nourish extremely low birth weight infants in
the NICU
BW < 1.5 kg Any BF on discharge
from hospital (not
specified further)
201 67% pre vs 70% post,
P = .74
Abbreviations: BF, breastfeeding; BW, birth weight; CGA, corrected gestational age; DHM, donor human milk; HMB, human milk bank; NEC, necrotizing enterocolitis; NICU, neonatal intensive care unit;
OMM, own mother’s milk.
6
Table 2. Quality Assessment.
Authors
Study Design
(Score)
Sample
Size (Score)
Quality of Control Group
(Score)
Calculation of
OR/RR (Score)
Confounding
Factors (Score)
Geographical
Spread (Score)
Score (Quality
of Study)
Beasmore et al41 Retrospective
observational (0)
122 (1) Demographic variables noted, no
differences (2)
Yes (2) None (2) Data from 1 unit
(0)
7 (moderate)
Bishop et al36 Retrospective
observational (0)
331 (1) Demographic variables noted, no
differences (2)
Yes (2) None (2) Data from 1 unit
(0)
7 (moderate)
Delfosse et al13 Retrospective
observational (0)
650 (2) No control group (0) No (0) DHM introduced as
part of bundle (1)
Data from 2
units (1)
4 (low)
Esquerra-Zwiers
et al15
Retrospective
observational (0)
265 (1) Demographic variables noted, no
differences (2)
Yes (2) No data (0) Data from 1 unit
(0)
5 (moderate)
Kok et al42 Retrospective
observational (0)
155 (1) No demographic variables
documented (0)
No (0) No data (0) Data from 1 unit
(0)
1 (low)
Marinelli et al37 Prospective
cohort study (1)
154 (1) Demographic variables noted,
significant differences (1)
Yes (2) None (2) Data from 1 unit
(0)
7 (moderate)
Montgomery
et al14
Prospective
interventional (2)
245 (1) Demographic variables noted,
significant differences (1)
Yes (2) DHM introduced as
part of bundle (1)
Data from 1 unit
(0)
7 (moderate)
Parker et al43 Retrospective
observational (0)
154 (1) Demographic variables noted, no
comment on whether significant
differences between groups (0)
Yes (2) None (2) Data from 1 unit
(0)
5 (moderate)
Utrera Torres
et al40
Retrospective
observational (0)
122 (1) Demographic variables
documented, significant
differences (1)
Yes (2) None (2) Data from 1 unit
(0)
6 (moderate)
Verd et al39 Retrospective
observational (0)
201 (1) Demographic variables
documented, no differences (2)
Yes (2) None (2) Data from 4
units (2)
9 (high)
Abbreviations: DHM, donor human milk; OR, odds ratio; RR, relative risk.
Williams et al 7
(Table 1). A single-center study judged to be of moderate
quality found that the introduction of DHM was associated
with a significant decrease in the percentage of feeds that
were OMM days 1 to 14 (P < .01) and days 1 to 28 (P = .04)
of life.38 One study examined the percentage of exclusive
OMM given until full feeds were established41 and found no
significant difference (P = .51) between the pre- and post-
DHM groups. One study examined the percentage of feeds
that contained > 50% OMM given to infants of up to 34
weeks corrected gestational age36 and again found no differ-
ence between the groups (P = .95). Two studies looked at the
percentage of OMM given (as volume) for days 1 to 14 13
and 1 to 28 of life,37 respectively, but did not calculate a sta-
tistical summary measure.
Discussion
Interest in the use of DHM has increased over the past
decade, manifest by a worldwide expansion in the number of
HMBs. Despite this, there remains a relative lack of high-
quality research into the impact of DHM on the recipient
neonatal population or its wider societal effects. Our system-
atic review of the use of DHM on maternal breastfeeding
rates confirmed this lack of high-quality data, identifying
only 10 studies that met the inclusion criteria. These studies
were geographically limited, available from only 4 countries,
and most of the included studies (6/10) were from the United
States. Four of the 10 included studies were conference
abstracts and were therefore not peer reviewed. Using a scor-
ing system to assess study quality, only 1 was judged to be of
high quality.
The available data demonstrate mixed effects on mea-
sures of maternal breastfeeding when DHM is introduced to
a neonatal unit. Relative risk calculations with aggregated
data from 4 studies did show a significant increase in any
breastfeeding on discharge after the introduction of DHM.
However, there appeared to be no effect on exclusive breast-
feeding on discharge or the exclusive administration of
OMM in the first 28 days of life after the introduction of
DHM. Even where DHM was introduced as part of a care
bundle (as it was in 2 of the included studies),13,14 in indi-
vidual centers, there appeared to be no significant increase in
measures of maternal breastfeeding. Conversely, 1 of the 10
studies showed a statistically significant decrease in the use
of OMM after the introduction of DHM.38 This was posited
by the authors to be due to the fact that the provision of DHM
was discouraging mothers from expressing breast milk.
However, the remainder of the available evidence does not
support the hypothesis that the introduction of DHM has an
adverse effect on breastfeeding rates in NICUs.
Some of the heterogeneity in results may reflect the fact
that DHM can be used in a variety of ways. One study
described DHM as a “bridge” to be used until a mother is
able to express enough milk for her preterm infant,13 whereas
others describe the rationale for DHM as being a way to
reduce the volume of formula feeds being given to preterm
infants14 or as a means to more rapidly introduce enteral
feeds.40 Given that DHM is introduced for a variety of rea-
sons and in a variety of ways (as part of package of measures,
by opening an HMB, or by replacing preterm formula in
feeding guidelines), it is perhaps not surprising that no con-
sistent effect is seen on measures of maternal breastfeeding.
Limitations
Inclusion criteria and definitions of outcomes varied between
the studies, precluding a formalized assessment of a risk of
bias using a funnel plot. Where aggregated relative risks were
calculated, study groups’ patient populations differed in terms
of birth weight and gestation, as well as whether they included
infants who had died in their denominator, although the num-
ber of these was small. Our data samples were small for each
variable, and the calculated intervals were wide, so that small
but important effects in either direction could not be excluded
for exclusive breastfeeding on discharge or use of OMM in the
first 28 days of life. For other outcomes, the heterogeneity of
study variables and patient populations limited the ability to
meta-analyze the data. We are unable to comment on whether
having consistent definitions of patient population and study
outcomes would have supported a positive effect of DHM on
other indicators of maternal breastfeeding success. However,
it is likely that the larger data sets permitted by consistent defi-
nitions would have allowed a more definitive answer to the
question of whether DHM affects these.
Eight of the 10 studies were retrospective, and there was a
high risk of bias, with only 1 study judged to be of high qual-
ity. We attempted to rule out publication bias by contacting
experts in the field to see whether substantial unpublished
databases existed on this topic and could not find evidence for
any. However, it remains possible that reports of trials with
negative findings have not entered peer-reviewed journals or
been accepted for conferences. In addition, we were unable to
obtain unpublished data from the included studies on breast-
feeding rates that may have influenced our results.
While we chose to concentrate on surrogate markers of
how much OMM was provided during admission and on dis-
charge, the introduction of DHM to a neonatal unit may
affect other important outcomes. These include rates of
OMM initiation, the duration of provision of OMM, the total
proportion of human milk (ie, OMM and DHM) given to
infants during their admission, the length of hospital admis-
sion, and practices related to the fortification of human milk.
Thus, the narrow focus of our research question may limit
the applicability of the findings of this systematic review.
Conclusion
In summary, the available data demonstrate positive effects on
some, but not all, measures of maternal breastfeeding rates
when DHM is introduced to a neonatal unit. There is also
8 Journal of Human Lactation
some evidence that in certain settings, rates might actually
decrease. However, overall there is probably sufficient data
available to reassure clinicians that the introduction of DHM
in itself is unlikely to adversely affect breastfeeding rates. If
the introduction of DHM is to be promoted as a cost-effective
way of promoting maternal breastfeeding, further well-
designed studies with standardized populations, consistent use
of DHM, measurable breastfeeding outcomes, and economic
evaluation may help to inform uniformity of practice. Ideally,
these could be integrated into large randomized controlled tri-
als looking at the effects of DHM on clinical variables such as
mortality, NEC, sepsis, and longer term health benefits.
Acknowledgments
We thank Marshall Dozier, Academic Liaison Librarian at the
University of Edinburgh, for her input into our search strategy. We
also thank Professor Paula Meier, Professor Richard Schanler,
Professor Maria Quigley, Professor Kathleen Marinelli, Professor
Karen Simmer, and an anonymous reviewer for their expert advice
on data (published and unpublished) that might meet the inclusion
criteria for this review.
Declaration of Conflicting Interests
The authors declared the following potential conflicts of interest
with respect to the research, authorship, and/or publication of this
article: Nicholas Embleton chairs a working group for the British
Association of Perinatal Medicin producing a framework for the
use of DHM in the United Kingdom, and Thomas Williams and
Judith Simpson are members of this group. Harish Nair has indi-
cated he has no potential conflicts of interest to disclose.
Funding[AQ: 6]
The authors disclosed receipt of the following financial support for
the research, authorship, and/or publication of this article: Funding
for travel allowing the participants to meet to plan this research was
provided by the British Association of Perinatal Medicine. Nicholas
Embleton declares that he has received research funding from man-
ufacturers of artificial milk formula and spoken at industry-spon-
sored events. The other authors have indicated they have no
financial relationships relevant to this article to disclose.
Supplementary Material
Supplementary material for this article is available online at http://
jhl.sagepub.com/supplemental.
References
1. Quigley M, McGuire W. Formula versus donor breast milk
for feeding preterm or low birth weight infants. Cochrane
Database Syst Rev. 2014;4:CD002971.
2. Patel RM, Kandefer S, Walsh MC, et al. Causes and timing of
death in extremely premature infants from 2000 through 2011.
N Engl J Med. 2015;372(4):331-340.
3. Strengthening human milk banking: a global implementation
framework. http://www.path.org/publications/files/MCHN_
strengthen_hmb_frame_May2014.pdf. Accessed February 3,
2015.
4. Bharadva K, Tiwari S, Mishra S, et al. Human milk banking
guidelines. Indian Pediatr. 2014;51(6):469-474. http://www.
ncbi.nlm.nih.gov/pubmed/24986283. Accessed February 3,
2015.
5. European Milk Bank Association (EMBA). http://www.euro-
peanmilkbanking.com/russia.html. Accessed February 3, 2015.
6. World Health Organization (WHO). Feeding of low-birth-
weight infants in low- and middle-income countries. http://
www.who.int/elena/titles/full_recommendations/feeding_lbw/
en/. Accessed February 3, 2015.
7. Section on Breastfeeding. Breastfeeding and the use of human
milk. Pediatrics. 2012;129(3):e827-e841.
8. Arslanoglu S, Corpeleijn W, Moro G, et al. Donor human milk
for preterm infants: current evidence and research directions.
J Pediatr Gastroenterol Nutr. 2013;57(4):535-542.
9. Perrine CG, Scanlon KS. Prevalence of use of human
milk in US advanced care neonatal units. Pediatrics.
2013;131(6):1066-1071.
10. Hagadorn JI, Brownell EA, Lussier MM, Parker MGK, Herson
VC. Variability of criteria for pasteurized donor human milk
use: a survey of U.S. neonatal ICU medical directors [pub-
lished online September 29, 2014]. JPEN J Parenter Enteral
Nutr.
11. Zipitis CS, Ward J, Bajaj R. Use of donor breast milk in neo-
natal units in the UK. Arch Dis Child Fetal Neonatal Ed.
2015;100:F279-F281.
12. Arslanoglu S, Moro GGE, Bellù R, et al. Presence of human
milk bank is associated with elevated rate of exclusive breast-
feeding in VLBW infants. J Perinat Med. 2013;41(2):129-131.
13. Delfosse NM, Ward L, Lagomarcino AJ, et al. Donor human
milk largely replaces formula-feeding of preterm infants in two
urban hospitals. J Perinatol. 2013;33(6):446-451.
14. Montgomery D, Schmutz N, Baer VL, et al. Effects of institut-
ing the “BEST Program” (Breast Milk Early Saves Trouble) in
a level III NICU. J Hum Lact. 2008;24(3):248-251.
15. Esquerra-Zwiers A, Rossman B, Patel A, Engstrom J, Janes J,
Meier P. “It’s somebody else’s milk”: dynamic decision-mak-
ing process used by mothers of VLBW infants who provide
consent for donor human milk feedings. In: ISRHML 2014.
Charleston, SC; 2014.[AQ: 7]
16. ClinicalTrials.gov. Donor human milk and neurodevelopmental
outcomes in very low birthweight (VLBW) infants. https://clinical-
trials.gov/ct2/show/NCT01232725. Accessed February 9, 2015.
17. Unger S, Gibbins S, Zupancic J, O’Connor DL. DoMINO:
donor milk for improved neurodevelopmental outcomes. BMC
Pediatr. 2014;14:123.
18. PRISMA statement website. http://www.prisma-statement.
org/. Accessed April 15, 2015.
19. US National Library of Medicine. PubMed. 2013. http://www.
ncbi.nlm.nih.gov/pubmed. Accessed January 23, 2013.
20. Embase. http://www.embase.com. Accessed January 23, 2013.
21. Global Health Archive. http://www.cabi.org/publishing-
products/online-information-resources/global-health-archive/.
Accessed November 6, 2014.
22. Ovid SP. https://ovidsp.ovid.com/. Accessed October 8, 2014.
23. The Cochrane Library. http://www.thecochranelibrary.com/
view/0/index.html. Accessed November 6, 2014.
24. CINAHL Complete. Full-text nursing journals. EBSCO.
http://www.ebscohost.com/nursing/products/cinahl-databases/
cinahl-complete. Accessed November 6, 2014.
Williams et al 9
25. Global Health Library. http://www.globalhealthlibrary.net/
php/index.php. Accessed November 6, 2014.
26. Current Controlled Trials. Search the metaRegister of clini-
cal trials. http://www.controlled-trials.com/mrct/. Accessed
November 6, 2014.
27. PROSPERO: International Prospective Register of Systematic
Reviews. http://www.crd.york.ac.uk/PROSPERO/. Accessed
December 24, 2014.
28. Google Scholar. http://scholar.google.co.uk/. Accessed
November 6, 2014.
29. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging
consensus on rating quality of evidence and strength of recom-
mendations. BMJ. 2008;336(7650):924-926.
30. Jackson S, Mathews KH, Pulanic D, et al. Risk factors for
severe acute lower respiratory infections in children: a system-
atic review and meta-analysis. Croat Med J. 2013;54(2):110-
121. http://www.pubmedcentral.nih.gov/articlerender.fcgi?arti
d=3641871&tool=pmcentrez&rendertype=abstract. Accessed
February 3, 2015.
31. Kirkwood B, Sterne J. Medical Statistics. 2nd ed. Oxford,
UK: Blackwell Publishing Ltd; 2003:153-156. https://scholar.
google.co.uk/scholar?q=medical+statistics+2003+blackwell+
kirkwood&btnG=&hl=en&as_sdt=0%2C5#0. Accessed April
18, 2015.
32. Ward L, Auer C, Smith C, et al. The human milk project: a
quality improvement initiative to increase human milk con-
sumption in very low birth weight infants. Breastfeed Med.
2012;7:234-240.
33. Blomquist HK. Supplementary feeding in the maternity
ward shortens the duration of breast feeding. Acta Paediatr.
1994;83(11):1122-1126.
34. Simmer K, Hartmann B. The knowns and unknowns of human
milk banking. Early Hum Dev. 2009;85(11):701-704.
35. Sullivan S, Schanler RJ, Kim JH, et al. An exclusively human
milk–based diet is associated with a lower rate of necrotizing
enterocolitis than a diet of human milk and bovine milk–based
products. J Pediatr. 2010;156(4):562-7.e1.
36. Bishop CE, Vasquez MM, Petershack JA, Blanco CL.
Pasteurized donor human milk for VLBW infants: the effect
on necrotizing enterocolitis and related factors. J Neonatal
Perinatal Med. 2010;3(2):87-93.
37. Marinelli KA, Lussier MM, Brownell E, Herson VC, Hagadorn
JI. The effect of a donor milk policy on the diet of very low
birth weight infants. J Hum Lact. 2014;30(3):310-316.
38. Esquerra-Zwiers A, Wicks J, Rogers L, et al. Impact of Donor
Human Milk in a High Mother’s Own Milk Feeding Neonatal
Intensive Care Unit. International Society of Research in
Human Milk and Lactation; 2014.[AQ: 8]
39. Verd S, Porta R, Botet F, et al. Hospital outcomes of extremely
low birth weight infants after introduction of donor milk
to supplement mother’s milk. Breastfeed Med. 2015;10(3):
150-155.
40. Utrera Torres MI, Medina López C, Vázquez Román S, et al.
Does opening a milk bank in a neonatal unit change infant
feeding practices? A before and after study. Int Breastfeed J.
2010;5:4.
41. Beasmore J, Rumley F, Odd D, Wach R. Easy access to pas-
teurised donor breast milk reduces use of formula milk in
infants with birth weight < 1500. Paper presented at: UKAMB
Annual Meeting; 2014.[AQ: 9]
42. Kok CH, Hartmann B, Simmer K, Chiffings D. Influence of
the donor human milk bank on feed type of infants born at less
than 30 weeks gestation at discharge from the neonatal unit.
J Paediatr Child Health. 2010:13.[AQ: 10]
43. Parker MGK, Burnham LA, Sanchez E, Braun N, Philipp B,
Merewood A. Implementation of a donor milk program is
associated with greater consumption of mother’s own milk in
a single, U.S., level 3 NICU. In: Pediatric Academic Societies
Abstract Archive. 2014. http://www.abstracts2view.com/pasall/
authorindex.php. Accessed xxx.[AQ: 11]
